Skip to content

daynaskully/usercount

 
 

Repository files navigation

Mastodon User Count Bot

A bot which counts users from all instances listed at [https://instances.mastodon.xyz] then posts statistics to Mastodon.

My copy is currently running at https://social.lou.lt/@mastodonusercount

This is a variation of the bot by @josefkenny: https://github.com/josefkenny/usercount

Dependencies

  • Python 2
  • gnuplot version 5 or greater
  • Mastodon.py: pip install Mastodon.py
  • Everything else at the top of usercount.py!

Usage:

  1. Edit config.txt to specify the hostname of the Mastodon instance you would like to get data from.
  2. Create a file called secrets.txt in the folder secrets/, as follows:
uc_client_id: <your client ID>
uc_client_secret: <your client secret>
uc_access_token: <your access token>

To get these values, create an account for your bot, then run this script:

from mastodon import Mastodon

# change this to the apprpriate instance, login and username
instance_url = "https://mastodon.social"
user_name = "[email protected]"
user_password = "123456"

Mastodon.create_app("My User Count", scopes=["read","write"],
   to_file="clientcred.txt", api_base_url=instance_url)

mastodon = Mastodon(client_id = "clientcred.txt", api_base_url = instance_url)
mastodon.log_in(
   user_name,
   user_password,
   scopes = ["read", "write"],
   to_file = "usercred.txt"
)

Your client id and secret are the two lines in clientcred.txt, your access token is the line in usercred.txt. (Yeah, I know I should have automated this step -- but hey, the above script is better than having to figure it out by yourself! ;) )

  1. Use your favourite scheduling method to set ./usercount.py to run regularly.

Call the script with the --no-upload argument if you don't want to upload anything.

Note: The script will fail to output a graph until you've collected data points that are actually different!

About

User statistics bot for Mastodon

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 57.0%
  • Gnuplot 41.4%
  • HTML 1.6%